Neural crest stem cells (NCSCs) are a transient multipotent embryonic cell population that represents a defining characteristic of vertebrates. The neural crest (NC) gives rise to many derivatives including the neurons and glia of the sensory and autonomic ganglia of the peripheral nervous system, enteric neurons and glia, melanocytes, and the cartilaginous, bony and connective tissue of the craniofacial skeleton, cephalic neuroendocrine organs, and some heart vessels.
We present evidence that neural crest (NC) competence can be acquired very early when human embryonic stem cells (hESCs) are selectively neuralized towards dorsal neuroepithelium in the absence of feeder cells in fully defined conditions. When hESC-derived neurospheres are plated on fibronectin, some cells emigrate onto the substrate. These early migratory Neural Crest Stem Cells (emNCSCs) uniformly upregulate Sox10 and vimentin, downregulate N-cadherin, and remodel F-actin, consistent with a transition from neuroepithelium to a mesenchymal NC cell. Over 13% of emNCSCs upregulate CD73, a marker of mesenchymal lineage characteristic of cephalic NC and connexin 43, found on early migratory NC cells. We demonstrated that emNCSCs give rise in vitro to all NC lineages, are multipotent on clonal level, and appropriately respond to developmental factors. We suggest that human emNCSC resemble cephalic NC described in model organisms. Ex vivo emNCSCs can differentiate into neurons in Ret.k- mouse embryonic gut tissue cultures and transplanted emNCSCs incorporate into NC-derived structures but not CNS tissues in chick embryos.
These findings will provide a framework for further studying early human NC development including the epithelial to mesenchymal transition during NC delamination.
Citation: Curchoe CL, Maurer J, McKeown SJ, Cattarossi G, Cimadamore F, et al. (2010) Early Acquisition of Neural Crest Competence During hESCs Neuralization. PLoS ONE 5(11): e13890. doi:10.1371/journal.pone.0013890
Editor: Joseph Najbauer, City of Hope National Medical Center, United States of America
Received: April 7, 2010; Accepted: September 23, 2010; Published: November 9, 2010
Copyright: © 2010 Curchoe et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Funded by California Institute for Regenerative Medicine Seed grant RS1-00466-1 to Alexey V. Terskikh. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
NCSCs have been well characterized in a number of model organisms , , , , including mouse, chicken, frog, and zebrafish , , , but little is known about the mechanisms of human NC specification, migration and differentiation.
In humans, the NC starts to migrate before neural tube closure, as early as embryonic stage 9, around the third to fourth week of pregnancy . Multiple pathologies such as peripheral neuropathies, skeletal and nervous system disorders and pigment disorders stem from aberrant NC specification, migration or differentiation , , . The derivation of human NCSCs from human embryonic stem cells (hESCs) will help to define the cellular and molecular mechanisms operating in human cells and facilitate the development of diagnostic and therapeutic strategies.
The manipulation of signaling molecules and pathways to direct ESC differentiation has been widely reported in the literature. Historically, the derivation of peripheral nervous system (PNS) cell types from mouse and primate ESCs has relied on co-culture with the mouse stromal line PA6 and late exposure to BMP4  or, in the case of hESCs, co-culture with PA6 . Similar to the work reported here, NC derivatives were obtained from bovine inner cell masses, after induction of differentiation by withdrawal of growth factors and supplementation with ascorbic acid . NCSCs derived from hESCs after extensive passages and/or using a combination of stromal induction and the addition of BMP2 have been reported , . More recently, the generation of a mixed population, depending on the cell densities, of central nervous system (CNS) neural progenitors and NC was also reported , . The obligatory co-culture with PA6 or MS5 and the generation of only small numbers of some NC lineages (sometimes requiring prospective isolation of ~0.1% of the cell population) are hurdles in all of the current differentiation protocols. Additionally, because NC competency is a transient phenomenon that appears to be established very early during gastrulation , the possibility exists that extensive in vitro amplification alters the properties of early NC cells.
We report that under conditions selectively promoting neuralization of hESCs towards dorsal neuroepithelial fate, NC competence is acquired much earlier than previously possible. Human ESC-derived NCSCs upregulate SoxE genes, Wnt and TGFβ signaling pathways associated with NC specification, maintenance, and migration , . Using Sox10, a transcription factor playing a critical role in NC development , we identified a distinct population of Sox10-positive cells migrating from adherent hESC-derived neurospheres. These emNCSCs can differentiate into all NC lineages including PNS neurons and glia, smooth muscle myocytes, chondrocytes and melanocytes and respond to the well-known signals that block NC specification in other model systems . Migratory NCSCs can colonize aganglionic embryonic gut cultures, where they differentiate into neurons. Finally, grafting into chick embryos demonstrates that hESC-derived emNCSCs contribute specifically to proper NC derivatives, differentiating into neurons and glia in the cranial ganglia, glia along nerves, mesenchyme and connective tissues in cranial regions.
Culture of human ES cells
An NIH-approved commercially available human ES cell line (H9, obtained from WiCell, Wisconsin) was used to generate neural precursors/putative NC cells in vitro. Undifferentiated hESCs were maintained on irradiated mouse feeder layers and Matrigel (final dilution1:30; BD Biosciences, cat # 356231) in Knockout DMEM (GibcoBRL/Invitrogen, cat #10829-018) with 20% Knockout Serum Replacement (GibcoBRL/Invitrogen, cat #10828-028), non-essential amino acids (100 µM final; GibcoBRL/Invitrogen, cat #11140-050), L-glutamine (final 1 mM; GibcoBRL/Invitrogen, cat #25030-081), β mercaptoethanol (100 µM final; GibcoBRL/Invitrogen, cat #21985-023), and supplemented with bFGF (final 8 ng/ml; Sigma, cat# F -0291-25UG). hESCs were passaged every 7-10 days by separating cells manually or using collagenase IV (final 1 mg/ml; GibcoBRL/Invitrogen, cat #17104-019) diluted in KO DMEM. Medium was changed every day and differentiating colonies were manually removed.
Generation of emNCSC from human ESCs
The differentiation protocol was previously described , except that day 5 spheres were plated on fibronectin. Briefly, collagenase IV passaged cells were resuspended in PBS and large clusters, not appropriate for neurosphere formation, were gravity pelleted and discarded. Clusters of hESCs that were left in suspension (about 50-100 cells/cluster) were transferred to a new tube, gravity pelleted and washed with PBS one additional time to remove all residual media and single cells. Cells were resuspended in neurosphere suspension (NS) medium: 1:1 ratio of DMEM/F-12 Glutamax (GibcoBRL/Invitrogen, cat# 10829-018) and Neurobasal media (GibcoBRL/Invitrogen, cat# 11140-050), BIT 9500 (final 10%; StemCell Technologies Inc, cat# 09500), L-glutamine (final 1 mM; GibcoBRL/Invitrogen, cat# 25030-081),50 U/ml penicillin, 50 µg/ml streptomycin (Invitrogen, cat#15070-063), ) and B27 (final 5%; GibcoBRL/Invitrogen, cat# 17504-044) supplemented with insulin (final 5 ng/ml; Sigma, cat# I4011), nicotinamide (final 5 mM; Sigma, cat# N0636) supplemented with bFGF (final 20 ng/ml; Chemicon, cat# GF003) and EGF (final 20 ng/ml; Sigma, cat# E9644). Neurospheres were cultured for 5 days in non-adherent polypropylene dishes (Ted Pella, Redding, CA), changing the media every 2 days. Day 5 neurospheres were plated on fibronectin-coated (final 1 mg/ml) culture dishes in NS medium. EmNCSCs were passaged every 3–4 days with Accutase.
Illumina Microarray and Q-PCR analysis
Total RNA was extracted using the RNeasy kit and 1 µg of total RNA was reverse transcribed using the Quantitect kit (Qiagen) according to the manufacturer's suggestions. About 500 ng total RNA per time point was used for cDNA synthesis followed by cRNA synthesis/amplification and labeling using the “Illumina RNA amplification” kit (Ambion). The labeled cRNA was hybridized to Human WG-6 sentrix BeadChip Arrays (Illumina) and scanned following the manufacturer's instructions. Analysis was performed using the algorithms included with the GeneSpring software. Gene expression changes greater than 2-fold; parametric test, variance equal, P value cut off = 0.05; Volcano statistic analysis GeneSpring software were considered significant. For Q-PCR, GAPDH was used for normalization and the data were analyzed using the standard curve method. Microarray and Q-PCR were processed by the Burnham Institute's gene analysis core facility. Q-PCR was performed with SyberGreen master mix (Invitrogen) according to the manufacturer's recommendation. Primer sequences are provided in Table 1. Q-PCR was perfomed as follows: Initial denaturation: 10 min at 95°C; 40 cycles of denaturation: 30 sec at 95°C, annealing: 1 min at 56°C, extension: 30 sec at 72°C; and one final cycle of denaturation for 1 min at 95°C, annealing for 30 sec at 65°C and final denaturation for 30 sec at 95°C. Microarray hybridizations were performed on Illumina arrays and analyzed with Genespring software. All data are MIAME compliant and that the raw data has been deposited in a MIAME compliant database (GEO, accession number #pending).
Table 1. Primers used for Q-PCR of PAX3.doi:10.1371/journal.pone.0013890.t001
Cells were fixed with 4% paraformaldehyde in PBS for 15 minutes at room temperature and rinsed 3 times with PBS. Following fixation, cells were blocked in 3% BSA or 10% goat serum for 10 minutes, primary antibodies (Table 2) diluted in 3% BSA/PBS or 10% goat serum/PBS were then applied at the appropriate concentration and incubated either at room temperature for 1 hour or overnight at 4°C. Cells were then washed 3 times for 3 minutes, and incubated at room temperature with an appropriate fluorochrome-conjugated secondary antibody (1:1000 diluted in 3% BSA/PBS). Cells were washed 3 times for 5 minutes and incubated for 10 minutes with Hoechst dye #33342 or Dapi to label cell nuclei.
Table 2. Antibody purchasing and dilution information.doi:10.1371/journal.pone.0013890.t002
Clonal expansion of emNCSCs and lineage analysis
Clonal analysis was performed as described by the Studer group. Briefly, cells were enzymatically dissociated after exposure to Accutase for 10 min. The cells were sedimented by centrifugationand the pellet was resuspended and filtered through 40 µm mesh. Filtered cells were counted and directly plated onto polyornithine (1.6 mg/ml ) + fibronectin (2 µg/ml) coated dishes at clonal densities (~10 cells/cm2). Cells were grown in NS medium for 15 days followed by mitogen (bFGF and EGF) withdrawal and culture in NS medium supplemented with BDNF (10 ng/ml), NGF (10 ng/ml), GDNF (10 ng/ml), 1 mM dibutyryl cAMP, CNTF (10 ng/ml) and neuregulin (20 ng/ml) for 15 days. Cell were fixed in 4% paraformaldehyde for 10 minutes at room temperature and incubated with primary antibody overnight at 4°C, followed by 1 hour incubation with the secondary antibodies at room temperature.
Addition of exogenous factors to manipulate NC specification or migration
BMPs, Wnts and retinoic acid have all been found to play roles in the specification and migration of NC in model organisms , , . Neurospheres were cultured (suspended for 5 days with 3 days of adherent culture) in NS medium in the presence of 100 ng/ml Noggin (a BMP antagonist) or a conditioned medium from DKK1-producing 293T cells. Adherent cultures were then stained with the NC markers Sox10, p75 or HNK-1.
In vitro derivation of NC-derived lineages and directed differentiations
For spontaneous differentiation, emNCSCs were maintained at confluency for up to 3 weeks in NS medium. Basic NS medium (NS medium without EGF, FGF, insulin or nicotinamide) was modified as previously described for directed differentiation , , , ; Briefly, for neuronal differentiation, 50 ng/ml BMP2 (R&D Biosystems), for neuroglial, 10% FCS/5 mM Forskolin (Sigma) and for smooth muscle myocytes 1 ng/ml TGFβ3 (R&D Biosystems) were supplemented to NS basic medium. To induce chondrogenic differentiation, 1×106 emNCSCs were centrifuged at 800×g for 3 min. Pelleted cells were fed with chondrogenic medium (Lonza) supplemented with 10 ng/ml TGFβ3 (USBiological). Medium was changed every third day. Cultures were fixed after 4 weeks and chondrogenic differentiation was assessed by immunohistochemistry (Alcian blue).
Prospective isolation by Fluorescence Activated Cell Sorting
At day 9 of emNCSC derivation the adherent neurospheres (composed of neuroepithelial rosettes) were mechanically removed and putative migrating emNCSCs were detached with Accutase and resuspended in 3% BSA/PBS for 20 minutes to block non-specific binding. Cells were then incubated with p75 (1:300), HNK-1 (1:300) or CD73 (1:500) for 30 minutes. Cells were washed in 3 ml of PBS and resuspended in 3% BSA/PBS with appropriate secondary antibodies (1:1000) for 30 minutes and washed with 3 mls of PBS before being resuspended in 1% BSA/PBS and incubated with PI (live/dead stain). Cells were sorted according to fluorescence (FACSVantageSE DiVa, BD Biosciences, San Jose) and data were analyzed with FlowJo (version 8) software.
Injection of emNCSCs into organotypic fetal gut cultures
The intestine was isolated from day 10.5-11 Ret.k- embryos. Mice carrying the Ret.k- mutation were previously characterized and analysis of homozygous embryos has demonstrated that the aganglionic phenotype is due to an early defect of the enteric NC . The method for organotypic embryonic gut cultures was previously described . Briefly, E10.5- 11 Ret.k- mouse embryos were obtained from timed pregnant heterozygous matings. The developing gut was dissected and placed across a V-shaped notch cut into Millipore paper (0.45 µm, Black gridded type HAGB, Millipore, Bedford MA). Each gut segment suspended on filter paper was placed into a Tersasaki well (60-well plate, Disposable Products, Technology Park, SA, Australia) overfilled with media to cover the entire gut segment and cultured in 5% CO2 at 37° for 7 days. EmNCSCs were injected into the intestinal lumen and walls using a fire-polished glass needle connected to a mouth-injection tube. On average, 1×105 cells were injected per gut piece. After 7 days, gut cultures were used for whole mount immunostaining and cryosectioning. Whole mount immunostaining for neurofilament was performed as previously described . The Ret.k mice were housed in accordance with NIH guidelines and the Burnham Institute animal house rules.
In ovo transplantation.
Fertile chicken eggs were incubated until somite stage 7 to 12. Eggs were windowed and embryos visualized using India Ink (1:10). Pelleted emNCSCs (~10,000 cells) were injected beside the cranial neural tube, in the pathway of migratory NC. Eggs were sealed and incubated for 1–5 days, at which time the embryos were fixed and cryosectioned . Sections were incubated overnight with antibodies raised against: human nuclei (mouse IgG1, Chemicon MAB1281, 1:250), early neuronal marker HuC/HuD (mouse IgG2b, Invitrogen A-21271, 1:250 ) and HNK-1 (mouse IgM, DSHB, 1:100) and smooth muscle actin (mouse IgG2a, Sigma A2547, 1:800). Sections were then washed and incubated in conjugated secondary goat anti-mouse antibodies (all from Invitrogen): IgG1 Alexa 488, IgG2a Alexa 594 (both 1:2000), IgG2b Alexa 633 and IgM Alexa 350 (both 1:1000). Sections were imaged using Zeiss Axioskop plus microscope with AxioVision 4.6 software.
Derivation of human emNCSC from hES cells
hESCs can be efficiently differentiated into uniform neural progenitors, as previously shown . Around day 5 of differentiation, rosette structures appeared within floating neurospheres that resemble primitive neuroepithelium and are reminiscent of neural tubes , . The cells were collected and the dorsal – ventral identity of the cells was investigated (Figure 1). At this time of differentiation, most cells were positive for the pan-neuroepithelium marker Pax6 as well as dorsal neuroepithelial markers Pax3 and Sox9 but negative for the ventral neural tube marker Nkx2.2. The Q-PCR analysis was consistent with the immunostaining data. These results suggest that current neuralization conditions (see methods and expression analysis below) favor the derivation of dorsal neuroepithelium, which is the region of the neural tube from which NC arises in vivo , .
Figure 1. Neuralized human ES cell-derived express markers of dorsal neuroepithelium.
hESCs were neuralized for 5 days in floating spheres, dissociated and cytospins were immunostained for Pax6 (A), Sox9 (B), Pax3 (C), and Nkx2.2 (D). (E) Quantification of the results in A–D. (F) Q-PCR analysis of Sox9 and NKX2.2 expression in day 5 spheres (black bars) vs hESC (white bars). All values normalized to 18S expression; * = p<0.05, ** = p<0.0001, one-tail unpaired t-test. Scale bars 50 µM.doi:10.1371/journal.pone.0013890.g001
Primary murine NC cells are typically derived by plating neural tube explants from embryonic day 8.5 on fibronectin (a substrate permissive of NC migration in vivo ) and recovering the migrating cells . Therefore, we plated day 5 neurospheres containing rosette structures on fibronectin and assayed NC marker expression after several days of migration. We found that 3 days of exposure to fibronectin resulted in a high number of migratory cells, which we called early migratory NCSCs. These cells express the NCSC markers Sox10, p75 and HNK-1 (Figure 2). The emNCSCs were uniformly positive for Sox10, while attached neurospheres composed of neuroepithelial rosettes (Figure 2D) were uniformly negative for Sox10 (Figure 2B, E). In contrast, both attached neurospheres and migrating cells were heterogeneously positive for the classical NC markers p75 and HNK-1 (Figure 2C, E, G). Because the adherent neurospheres resemble dorsal neuroepithelium, this observation appears similar to the in vivo situation. Indeed, in mouse embryos (E9.5-E11) at least some cells in the dorsal neuroepithelium (premigratory NC) are positive for the p75 marker  and human neural tube explants were reported to express p75 . After an initial wave of NC emigration the adherent neurospheres (uniformly negative for Sox10) were scraped off the plate and replated. A second wave of emigrating cells was observed, which also acquired the Sox10 and p75 markers (Figure 2G–H). Most cells within the replated neurospheres remained Sox10-negative and the secondary wave of emigrating cells appeared to have higher levels of the p75 marker. These results suggest that many, if not all, cells within the adherent spheres are competent to upregulate Sox10 upon delaminating from neuroepithelial clusters and initiating migration on fibronectin. This behavior is reminiscent of an epithelial to mesenchymal transition (EMT), a characteristic feature of NC cells in model organisms .
Figure 2. Early migratory NCSCs upregulate Sox10.
Immunostaining of adherent neurospheres and emigrating NCSCs after 3 days on fibronectin-coated plates. (A) DAPI staining for nuclei. Note the rosette structures within the adherent neurospheres. (B) Sox10 staining is observed in emigrating cells, but not in the adherent neurospheres. (C) p75 staining is observed in adherent neurospheres and emigrating cells. (D) High magnification of the overlay in B–C. (E) Confocal image of emigrating NCSCs confirms the membrane localization of p75. (F) HNK-1 staining is observed in adherent neurospheres and emigrating cells. (G, H) Adherent neurospheres were manually removed and replated on fibronectin and the second wave of emigrating cells was observed. (G) Secondary emigrating cells acquire Sox10, while the majority of cells in the replated neurosphere are Sox10-negative. (H) The expression of in the second wave of emigrating cells. Scale bars in A, B, C and D represent 100 µm and 10 µm in E Bars in F,G and H depict 150 µm.doi:10.1371/journal.pone.0013890.g002
Emigration of emNCSCs from neuroepithelial clusters resembles EMT
Adherent neurospheres display thick bands of f-actin at the cell borders, characteristic of epithelial cells. Cells migrating from neurospheres acquire vimentin expression (Figure 3C–D) and have thin bands of f-actin at the cell borders and diffuse f-actin expression throughout the cell body (Figure 3A–B), consistent with observations in mesenchymal cells . This alteration of the f-actin distribution between the rosette cells and the migrating cells is reminiscent of the EMT observed in NC cells delaminating from the neural tube in vivo . Similarly, the adherent neurospheres are positive for N-cadherin and emNCSCs downregulate N-cadherin expression, in further support of an EMT . In addition, emNCSCs upregulate the expression of gap junction protein connexin 43 (Cx43) at the onset of migration (Figure 3F). The expression of Cx43 was found to be critical for NC cell survival and characteristic of early migratory NC cells in vivo . In summary, the transition from adherent neurospheres (further referred to as neuroepithelial clusters) to migratory Sox10-positive cells has many features of classical EMT observed during delamination of NC in vivo in model organisms.
Figure 3. Epithelial to mesenchymal transition in migratory emNCSCs.
(A) Phalloidin stains thick bands of f-actin within the adherent Sox10-negative neurosphere and thin cell border bands. Diffuse f-actin staining is visible in migratory Sox10-positive cells. (B) Connexin 43 (red) is acquired in migratory emNCSCs. Attached sphere is outlined. (C) N-Cadherin (green) is present in adherent neurospheres and gradually downregulated in migratory emNCSCs visualized by the nuclear DAPI IHC. (D) Confocal image of thick f-actin (red) bands at the border of an adherent neurosphere (outlined). Vimentin (green) is acquired in emNCSCs migrating away from the attached neurosphere. Scal bars in A and B represent 50 µm, in C 25 µm, and in D 10 µm. Cell nuclei were visualized by DAPI unless indicated differently.doi:10.1371/journal.pone.0013890.g003
Microarray analysis of neuroepithelial clusters and migratory NCSC
The above results provide some distinctions between the Sox10-positive cells in neuroepithelial clusters and the Sox10-positive migratory cells. To further investigate the molecular difference between the two populations, we mechanically separated adherent neurospheres from migratory cells and performed microarray analysis of gene expression (Figure 4). We found that in migratory cells, 84 genes were significantly upregulated and 103 were downregulated compared to the neuroepithelial clusters. Confirming the immunostaining data, the SOX10 transcript was upregulated in migratory cells. Several well-known regulators of the NC lineage were selectively upregulated in migratory cells (e.g., PhOX2b, RGS 4, HEY1, DLK1, and GnRH2). Several other NC genes such as FOXD3, MSX1, PITX1, and Hand1 were upregulated in migratory cells. Curiously, IGFBP5, a negative regulator of NC and a potential regulator of craniofacial skeletogenesis , was selectively downregulated in migratory cells, suggesting a potential interaction between the neuroepithelial clusters and emigrating cells.
Figure 4. Microarray analysis of the adherent neurospheres and emNCSCs.
The top most upregulated gene transcripts in emNCSCs (red) and neuroepithelial clusters (blue) are shown. The middle insert highlights the selected set of NC-related genes upregulated in emNCSCs.doi:10.1371/journal.pone.0013890.g004
In addition to the expression analysis described above, we determined the dynamic expression of key NC-related genes over 10 days of neuralization using the entire cultures (Figure S1). Among the upregulated transcripts were the growth factors Wnt3a (2-fold at 72 hours), Wnt1 (2-fold at 96 hours) and BMP4 (2-fold by day 3) as well as the downstream transcription factors Sox9 (2-fold at day 3), Sox10 (7-fold day 6), ZIC1 (2-fold at day 3), MSX1 (2-fold by day 2) and SNAIL, (2-fold by 60 hours). In addition, we used Q-PCR to examine the expression of Pax3, one of the earliest gene defining the NC field . PAX3 transcripts were found to be upregulated 2-fold at day 3 and 10-fold at day 5 of hESC neuralization.
We demonstrated a dynamic regulation of most NC-related genes during hESC differentiation towards emNCSC, including the upregulation of many NC-related genes in emNCSC when compared to the neuroepithelial clusters. These data contribute to our understanding of gene regulatory network orchestrating neural crest formation and provide support to our interpretation of the in vitro human model system - the cells within the clusters are similar to the dorsal neuroepithelium, while emigrating cells resemble NCSCs.
p75 expression does not discriminate emNCSCs from neuroepithelial clusters
Previously, the p75 marker was used for prospective isolation of NCSCs from rat embryo sciatic nerves  and when differentiated from hESCs after prolonged differentiation on stromal feeders . We found that the membrane-specific immunostaining for p75 was present in both clusters and emigrating cells (Figure 2C–E). We employed flow cytometry to prospectively isolate cells from hESC neuralization cultures at day 8, based on p75 expression. Consistent with the immunostaining data, most cells were positive for p75 (Figure 5A). Cells expressing high, medium, and low levels of p75 were isolated by FACS and their Sox10 levels and differentiation potential were examined (Figure 5). Most cells, whether p75 high, medium, or low, expressed Sox10 (85%, 80%, and 90% respectively, Figure 5B). After one week of spontaneous differentiation on fibronectin young neurons (βIII-tubulin), glia (GFAP), smooth muscle cells (SMA), as well as undifferentiated cells (nestin) were found in all cultures (Figure 5C and data not shown). No selectivity for differentiation into any particular lineage could be detected in the three populations. Another marker, HNK-1, was previously proposed as a marker of NC in chick embryos  and was used in the hESC system . Flow cytometry analysis of day 8 cultures revealed that over 90% of cells are positive for the HNK-1 marker (data not shown). These results suggest that the p75 and HNK-1 markers, previously used to prospectively isolate human NCSCs after prolonged differentiation on stromal feeders  are ineffective in discriminating migratory cells from the neuroepithelial clusters in our cultures. Because many (if not all) cells in the clusters can give rise to the migratory Sox10+ cells (e.g. when re-plated on fibronectin) it is possible that these pre-migratory NC cells already upregulated the p75 and HNK1 molecules.
Figure 5. p75 expression doesn't discriminate emNCSCs from neuroepithelial clusters.
(A) Day 8 cultures (containing emNCSC and neuroepithelial clusters) can be FACS-sorted based on p75 expression defining three populations: negative, medium and highly positive for p75. The grey profile corresponds to control situation with the secondary antibody only. (B) All fractions of cells were positive for Sox10 (85, 70 and 60% respectively for high, medium and negative). (C) All fractions can differentiate into neuronal, glial and smooth muscle lineages (p75-negative fraction is shown as an example). p75-negative sorted cells are reactive for; I. GFAP, II. β-3-tubulin, III. nestin, IV. smooth muscle actin. Cell nuclei are visualized by DAPI. Magnifications are 100x. Scale bars 30 µM.doi:10.1371/journal.pone.0013890.g005
Differentiation of emNCSC into the NC lineages
We determined whether emNCSCs as a population were able to give rise to a variety of NC lineages found in vivo, including neurons, glial or mesenchymal lineages, melanocytes, and adipocites. EmNCSCs were allowed to spontaneously differentiate in dense cultures through contact inhibition and were assayed for NC derivatives after 2–3 weeks (Figure 6 and Figure S2). After ~2 weeks in culture, peripheral autonomic lineages were identified by co-staining of TH and peripherin  (Figure 6A) and sensory peripheral lineages were detected by BRN3a (premigratory crest and neurons of the DRG) reactivity (Figure 6B). Neuronal derivatives were found to comprise 30% ±10% of derivatives. Myocytes were identified by smooth muscle actin staining after one week in culture (Figure 6C), and were also observed to comprise approximately 30% ±10% of derivatives. Consistent with events during normal developmental and in agreement with the previous observations , neurogenesis preceded gliogenesis. Characteristic markers of Schwann cells could be detected in cultures after emNCSCs were passaged for 3–4 weeks and differentiated for additional two weeks (Figure 6D). Glial derivatives comprised 20% ±10% of total derivatives. Melanocytes (comprising 5% ±3% of derivatives) were observed after 3-4 weeks in culture. Melanocytes were identified by their characteristic black melanosome inclusions and Melan A reactivity (Figure 6E).
Figure 6. Spontaneous differentiation of emNCSCs.
(A) Peripheral neuron-like cells are co-stained with TH (green) and Peripherin (red). Individual channels are shown in the inserts.(B) Smooth muscle cells reactive to smooth muscle actin. (C) The addition of 1% horse serum to the NS medium yielded GFAP-positive cells co-expressing p75NTR and proteolipid protein (PLP), consistent with Schwann cell identity. Inset, PCR detection of GFAP transcript. CTRL = NS medium. Scale bar = 25 ìm. (D) Brn3a (red) and Peripherin (green) staining indicates differentiation into peripheral sensory-like neurons. (E) Bright field image of black inclusion bodies in a characteristic melanosome, inset box shows Melan A (green) reactivity. Nuclei were visualized by DAPI. Scale bars represent 40 µm in A, C and D, 60 µm in D, and 100 µm in E and its inset.doi:10.1371/journal.pone.0013890.g006
Previously, NCSCs were observed to respond to various combinations of factors, selectively generating specific NC lineages , , . The addition of single growth factors to NS basal medium (no growth factors) led to enrichment of distinct cell types in the cultures. Addition of TGFβ initiated mainly two cell types: neurons and myocytes at a ratio of 70% ±5%: 28% ±4% (Figure 7A–D, TGFβ). To enrich for neuro-glial precursors, 10% fetal calf serum and forskolin (5 µM) were added to NS base medium. Under these conditions, glia comprised 30% ±2% of the cells, neuronal markers were detected in 60% ±5% of the cultures and no SMA-positive cells were detected (Figure 7E–H, FCS/Forskolin). Culturing emNCSCs with BMP2 enriched for a neuronal phenotype (Figure 7I–L, BMP2). Fifty percent ±7% of cells were positive for β III tubulin. Five percent ±2% were positive for GFAP, indicating glial differentiation and another 40% ±3% stained positive for SMA, indicative of myocyte differentiation. For chondrocytic differentiation, emNCSCs were incubated in defined medium, in a high-density culture, under low-oxygen conditions for 4 weeks . Comparable to data published earlier, this method resulted in cell aggregates that when sectioned, showed strong expression of collagen in the intercellular space as shown by Alcian blue staining (Figure 7M and N).
Figure 7. Directed differentiation of emNCSCs.
The growth factors FGF, EGF, insulin, and nicotinamide were withdrawn from NS medium and the base medium is supplemented with TGFβ (A-D) to enrich differentiating cells for neuronal fates, FCS/Forskolin to enrich for neuro-glial fates (E–H) and BMP2 to enrich for smooth muscle (I–L). A, E and I show GFAP immunohistochemistry (green, astrocytes). B, F and J indicate bIII tubulin (green, immature neurons). C, G and K show nestin (red, neuroblasts) and D, H and L show SMA reactivity (green, smooth muscle actin). Cell nuclei were visualized by DAPI, bars in A–L represent 100 µm. (M,N) Chondrocytic differentiation can be directed in emNCSC cultures (see materials and methods for details) and cryosectioned pellets reveal collagen II by Alcian blue staining. Scale bars 100 µM in M and 25 µM in N.doi:10.1371/journal.pone.0013890.g007
In summary, spontaneous and directed differentiation of human emNCSCs in vitro into multiple NC-specific lineages demonstrates their broad potential and ability to respond appropriately to developmental cues.
Clonal expansion of emNCSCs and lineage analysis
Experiments in model organisms suggested that in vivo, NC cells include multipotent, oligopotent and unipotent clones , . In order to assess the differentiation potential of emNCSCs at a single cell level, we performed clonal analysis experiments by plating single cell dissociated emNCSCs at clonal densities as previously described by the Studer group. Staining of 20 clones for markers of neurons (TuJ1), glial cells (GFAP) and smooth muscle cells (SMA) revealed 12 clones (63%) positive for all three lineage markers, 3 clones (16%) positive for TuJ1 and GFAP, 3 clones (16%) positive for SMA only, and 1 clone (5%) positive for TuJ1 only (Figure 8). These results suggest that emNCSCs are multipotent on the clonal level.
Figure 8. Clonal analysis of emNCSC.
Representative examples of multipotent (A–C), bi-potent (F), and unipotent (D, E) clones of emNCSC. For all images: TuJ1 (green), SMA (red), GFAP (blue); Scale bars 50 µM.doi:10.1371/journal.pone.0013890.g008
Inhibition of BMP and Wnt signaling re-specifies emNCSCs into CNS cells
Exogenous factors, such as BMPs and Wnts are well-known modulators of NC specification and migration in vivo , , . BMP4 is expressed in the avian dorsal neural tube at the time of neural tube closure when NC migration starts. Addition of Noggin at this stage, but not at later stages, inhibits NC formation and migration . We tested the effect of Noggin on the derivation of emNCSCs from hESCs. Control cells show Sox10, HNK-1 and p75 reactivity (Figure 9A–C). The addition of Noggin (100 ng/ml, starting on day 0, medium was changed every other day) nearly ablated expression of the NC markers Sox10, p75 and HNK-1 (Figure 9D–F). At the same time, the migration of cells from each adherent neurosphere was less than in controls (70% of control, Figure 9J). Similarly, the addition of the Wnt/β-catenin antagonist Dkk1 ablated expression of the NC markers Sox10 and HNK-1, but not the low affinity neurotrophin receptor p75. The retention of p75 is consistent with the finding that Wnt signaling does not change the axial identity of cells in the developing neural tube, but rather re-specifies the cell fates at the early stages of NC development . These results indicate that human emNCSCs appropriately respond to the stage-specific developmental signals, consistent with the early migratory nature of the emNCSCs identified in this study.
Figure 9. NC competency is ablated by treatment with Noggin or DKK.
The effect of Noggin (BMP inhibitor) and DKK (Wnt inhibitor) on hESC-derived emNCSCs. Untreated (control) neurosphere cultures are strongly reactive for (A) SOX10 (green), (B) HNK-1 (green), and (C) p75 (red) and negative for the CNS marker (D) Olig2 (red). The addition of Noggin to NS medium ablates (E) SOX10 (green), (F) HNK-1 (green), and (G) p75 (red) and induces expression of the CNS marker (H) Olig2 (red). Culturing neurospheres from day 0 in media that was collected from 293T cells transfected with a DKK plasmid ablates expression of the NC markers (I) SOX10 and (J) HNK-1 while the low affinity neutrophin receptor (K) p75 was still expressed. The insets in upper left corners show the cell nuclei (DAPI). Scale bars represent 50 µm in A, and C–H, 10 µm in B, and 100 µm in I–K.doi:10.1371/journal.pone.0013890.g009
Mesenchymal potential of emNCSCs
In vivo, early migratory NC cells (i.e. cephalic NC) are considered a uniform population of cells, capable of differentiation into mesenchymal lineage (e.g., myocytes) and peripheral neurons and glia , . In a previous study, human ESC-derived NCSCs were found to contain a small population of mesenchymal precursors, identified by the CD73 marker . We examined CD73 expression in day 5 neurospheres prior to culturing on fibronectin, as well as in epithelial clusters and migratory cells 3 days post plating on fibronectin. The day 5 neurospheres displayed no CD73 as expected for undifferentiated cells of neural origin (Figure 10A). At day 8 post plating, neuroepithelial clusters were also negative for CD73, however, ~13% of the migrating emNCSCs were positive for CD73 (Figure 10B–C). To investigate their differentiation potential, CD73-positive and -negative populations were isolated by FACS. Both populations were allowed to differentiate for one week after sorting and assayed for SMA, nestin, MAP2 and Ki67 markers. Both CD73-positive and -negative sorted cells gave rise to abundant SMA+ and nestin+ cells. GFAP+ cells were rare in both cases; however, CD73-negative cells produced approximately 10-fold more MAP2-positive cells than the CD73-positive fraction (Figure 10D). These data suggest that CD73-positive cells are much less efficient at differentiation into the neuronal lineage, as expected for the mesenchymal component of NC. Further experiments are needed, however, to demonstrate the selective advantage CD73-postive cells for mesenchymal differentiation.
Figure 10. CD73 expression in emNCSCs.
Early migratory NC cells are considered to be a uniform population of cells capable of differentiation into mesenchymal and PNS derivatives. To evaluate the potential of hESC-derived emNCSCs the mesenchymal marker CD73 was investigated at several time points, followed by prospective isolation and differentiation. (A) Day 5 neurospheres were tested for CD73 expression, approximately 0.11% of cells were positive. (B) Adherent neuroepithelial clusters were mechanically separated from migrating cells and tested for CD73 expression, approximately 0.5% of adherent neuroepithelial rosettes cells were positive. (C) 13% of cells migrating from neuroepithelial rosettes (SOX10-positive cells) are positive for CD73. (D) Migrating cells were sorted based on CD73 expression and cultured for 1 week. Both CD73-positive and -negative cells yielded nestin- and SMA-positive cells (areas with abundant staining are shown), however, the MAP2 reactive cells were enriched 10-fold in the CD73-negative fraction. Both populations had rare GFAP-positive cells and scattered KI67-positive cells. Scale bars represent 10 µm in I, I′ and V′, 50 µm in III, IV, III′ and IV′, and 100 µm in II, V and II′.doi:10.1371/journal.pone.0013890.g010
Injection of emNCSCs into organotypic embryonic gut cultures
The ability of emNCSCs to colonize and differentiate into appropriate NC derivatives in a disease model was tested by injection of emNCSCs into aganglionic embryonic gut cultures derived from mice carrying Ret.k- mutation. As previously described, only NC-derived enteric neurons can repopulate the Ret.k- intestine . Undifferentiated emNCSCs derived from hESC engineered to express GFP were injected into the knockout gut segments. After 7 days of culture, the tissue was cryopreserved, sectioned and immunostained for GFP and neurofilament. emNCSCs integrated into the Ret.k- intestinal wall and differentiated into neurons within the 7 days of culturing (Figure 11A). Because Ret.k- embryos totally lack the enteric nervous system in the small and large intestine , neurofilament-positive cells must be derived from injected human emNCSCs. Indeed, no staining for neurofilament was detected in the non-injected gut segments (Figure 11B). The endogenous enteric neurons were detected in the heterozygous animals (Figure 11C). The Ret.k- animals are unable to generate enteric neurons . However, the Ret.k- intestines injected with human emNCSC show immunostaining for both GFP and neurofilament, likely due to the generation of neurofilament+ cells from human emNCSC (Figure 11D). These data suggest that emNCSCs are able to colonizing the Ret.k- gut and differentiating into neurofilament-positive cells (presumably enteric neurons).
Figure 11. Injection of emNCSCs in aganglionic bowels in ex vivo organotypic cultures.
EmNCSCs can colonize aganglionic guts in organotypic gut cultures and differentiate into neurons. (A) Confocal image of a cryosectioned Ret KO gut segment 7 days post injection labeled with fluorescent antibodies against neurofilament (red). (B) Non-injected Ret KO gut segment 7 days post injection showing no neurofilament (red) reactivity. (C) In heterozygous animals endogenous neurons are present in the non-injected gut cultures (neurofilament staining in red). Cell nuclei are visualized by DAPI. Scale bars in A-C,10 µm. (D) Whole mount of a Ret KO gut segment stained with a neurofilament marker (red) and anti-GFP expressed by emNCSC (green). The GFP staining overlaps with the neurofilament staining in several places showing emNCSC-derived neurons; one example is marked by asterisk. Single channels (red, neurofilament and green, GFP) are depicted on top of the merged photograph for better identification (Magnification 20x).doi:10.1371/journal.pone.0013890.g011
In ovo transplantation of emNCSCs
Finally, we investigated the developmental potential of emNCSCs, namely, their ability to migrate, integrate and differentiate in vivo. GFP-labeled emNCSCs were injected into the chick embryo NC migration pathway beside the midbrain and rhombomeres 1 and 2 during the endogenous cranial NC migration (Figure 12A), and assayed for migration, appropriate targeting and differentiation. Cells were identified using immunohistochemistry of human nuclear antigen and markers for terminal differentiation. Five days after injection, transplanted cells were found in many of the destinations of cranial NC such as the branchial arches, the cranial ganglia and the head mesenchyme. Transplanted cells that migrated from the graft into the mesenchyme lost expression of HNK-1 (Figure 12B) and contributed to a variety of NC-derived tissues, both neural and mesenchymal. Some transplanted cells remained at the graft site and expressed low levels of HNK-1 but no neuronal markers (Figure 12B). Transplanted cells located throughout the cranial mesenchyme and along skeletal muscle fibers have elongated nuclei and expressed HNK-1, in this context, characteristic of Schwann cells (Figure 12C, E, H). Some transplanted cells were observed in undifferentiated cranial mesenchyme (Figure 12A, G), and among skeletal muscle fibers, potentially forming connective tissue. Many transplanted cells were located around blood vessels, some of which were positive for SMA (Figure 12D). emNCSCs were also seen differentiating into smooth muscle in other appropriate locations (Figure 12H). Numerous transplanted cells were found in the trigeminal and ciliary ganglia (Figure 12B and F), some of which expressed neuronal markers (Figure 12F) while others displayed glial morphology.
Figure 12. Injection of emNCSCs into the cranial NC pathway of chick embryos.
(A) Whole mount view of a clump of grafted emNCSCs (arrow) injected into the cranial region of a stage 11 chick embryo. (B–H) Sections through the cranial region of chick embryos 5 days post-injection labeled with fluorescent antibodies to human nuclei (HuN, green), HNK-1 (blue), neuronal marker HuC/HuD (purple) and smooth muscle actin (SMA, red). (B) A number of human emNCSCs remain clustered and express HNK-1 (arrows) while individual cells that have left the graft site (arrowheads) are negative for HNK-1 and contribute to head mesenchyme. (C) emNCSCs along HNK-1-positive nerves (arrowheads). (D) SMA-positive (arrow) and -negative (arrowhead) emNCSCs around a blood vessel. (E) HNK-1-positive emNCSCs within skeletal muscle (arrowhead) do not express muscle actin (red). (F) Some emNCSCs (arrows, left and right panels) in the TG express the neuronal marker HuC/HuD (middle and right panels). (G) emNCSCs in undifferentiated mesenchyme near blood vessels around the perimeter of the midbrain. No emNCSCs were observed in the midbrain. (H) Differentiated emNCSCs (left and right panels) expressing SMA (arrow, middle and right panels) or HNK-1 (arrowhead, middle and right panels). BV - blood vessel; TG – trigeminal ganglion. Scale bars 200 µM in A, 100 µM in B, 20 µM in C–H.doi:10.1371/journal.pone.0013890.g012
Both the location and the differentiation fates of transplanted cells were appropriate for cells of NC origin. In contrast, we failed to observe emNCSCs differentiating into cell types not normally derived from endogenous NC, such as neurons in the CNS. We concluded that emNCSCs transplanted in chick embryos during cranial NC migration respond to developmental cues, migrate along endogenous NC cells and differentiate into the appropriate tissues/structures.
Here, we characterize a population of cells emigrating from hESC-derived adherent neurospheres plated on fibronectin and provide evidence that they mimic the dorsal neural tube containing premigratory NCSCs and resembling their site of origin in vivo. The adherent neurospheres are mainly composed of neuroepithelial rosettes, which express Pax3 and Sox9, markers of dorsal neuroepithelium and premigratory NC , , , but are negative for the ventral neural tube marker Nkx2.2 . Cells that leave neuroepithelial clusters and start migrating on fibronectin uniformly upregulate Sox10, a critical transcription factor for NC development  (Figure 13). The clusters remain negative for Sox10 and can be mechanically isolated. After replating on fibronectin, they give rise to a second wave of emigrating cells, which also upregulate Sox10. Most of the cells within replated clusters remain negative for Sox10 3 days after replating. It is likely that virtually all cells within the original adherent neurospheres are competent to emigrate and upregulate Sox10. Indeed, over 80% of cells within adherent neurospheres are positive for the markers of dorsal neuroepithelium and lack the ventral markers, consistent with the cells being premigratory NC.
Figure 13. Scheme of early emNCSCs derivation from hESCs and differentiation into NC lineages.
hESC clusters are suspended for 5 days in NS medium. On day 5 of suspension culture neurospheres were allowed to adhere on fibronectin-coated plates and cultured in NS medium for 3 more days. Adherent neurospheres composed of neuroepithelial rosettes are negative for Sox10 and CD73. Migrating cells are positive for the NC markers Sox10 and CD73 (13%). Both cell types are positive for p75 and HNK-1. Rosettes can be manually excised and replated (dotted arrow) to yield additional waves of Sox10-positive emNCSCs. The adherent rosettes are therefore competent to become NC after they have undergone the EMT. Migratory emNCSCs can be allowed to spontaneously acquire NC lineage fates or can be directed to enrich for certain cell fates by the addition of growth factors.doi:10.1371/journal.pone.0013890.g013
Analysis of global gene expression profile of cells within the clusters and emNCSCs revealed selective upregulation of the NC-related genes in the migratory cells. For example, homeobox gene PhOX2b is essential for the development of both autonomic and sympathetic neurons ,  and regulates the dynamic expression of RGS4 . The upregulation of DLK1 and HEY1 in emNCSCs is consistent with the role of Notch signaling in NC differentiation and maintenance . We detected selective upregulation of gonadotropin-releasing hormone 2 (GnRH2) in emNCSCs. Indeed, some cells releasing GnRH arise from cranial NC  and their differentiation is modulated by the NC transcription factors SOX10 and FOXD3 . It is tempting to speculate that these differences between the migratory cells and adherent clusters will provide insight into the genes and pathways involved in delamination of human NC.
In contrast to previous reports , a much larger fraction (~100-fold increase) of emNCSCs is positive for the mesenchymal cell marker CD73, which is a unique characteristic of cephalic NC. Cephalic NC cells are known to emigrate earlier than trunk NC in all species studied, including recently derived human NC cells . Previously described NCSCs lack the expression of connexin 43 ( supplementary microarray data), which could be due to the extensive passaging. In contrast, the emNCSCs described in this study uniformly express connexin 43, previously found to be characteristic of early migrating NC . Finally, previous studies in chick demonstrated that BMP inhibition (e.g., with Noggin) can block NC specification during early, but not late stages of NC development . Treatment with Noggin quantitatively abolishes the appearance of emNCSCs. Taken together these data strongly argue that we have identified and characterized early migratory NCSCs, likely resembling cephalic NCSCs, derived from human ES cells.
EmNCSCs express the classical marker of NCSC, p75; however, we found that in our setting, dorsal neuroepithelium-like cells also express p75. Similarly, both cell types are positive for HNK-1, a well-known NC marker in avian but not murine embryos . Indeed, at the time of NC emigration in mice the most dorsal part of the neural tube, including some premigratory NC cells residing in that area, are positive for p75 . Similarly, human neural tube cells express p75 . Previously, when hESCs were neuralized using stromal feeder cells , , only a proportion of cells in culture (e.g., the most outer part of neuroepithelieal rosettes) were found positive for both p75 and HNK-1 . The difference in the neuralization protocols is a likely explanation for the observed discrepancies. The heterogeneous p75 reactivity within neuroepithelial clusters was observed using three different antibodies; rabbit polyclonal serum and two mouse monoclonal antibodies (including one directly conjugated, previously used for isolation of human NCSCs .) For HNK-1 staining, we used the same monoclonal antibody staining conditions as previously reported .
Recent evidence in avian embryos suggests that the onset of NC specification occurs during gastrulation, where epiblast cells that express the transcription factor Pax7, are specified to form NC cells in the absence of exogenous factors but in a manner that requires Pax7 itself . In the mouse NC, the Pax3 protein may be functionally equivalent to the Pax7 . We observed early and robust upregulation of Pax3 during hESC neutralization, such that by day 5 of neuralization ~80% of cells were positive for Pax3. Remarkably nearly 60% of the cells were positive for Sox9, a marker of premigratory NCSCs. These findings suggest that under current neuralization conditions most neural progenitors express markers associated with NC competence and specification. Despite their early specification at gastrula stages, NC cells do not delaminate from the neural tube until late neurulation and do not initiate differentiation until late migratory and post-migratory stages. Furthermore, execution of NC cell fates in vivo requires continuous Wnt signaling  while various growth factors appear to direct differentiation along distinct pathways , . Wnts and BMPs mediate NC induction in several model species , , . In a previous study using human cells, melanocytes were generated from hESCs using Wnt3a-conditioned media; however the presence of multipotent NCSCs was not demonstrated . Indeed, microarray analysis of neurosphere cultures confirmed the upregulation of Wnt1, Wnt3a, BMP4 and BMP5, as well as the expression of transcription factors downstream of Wnt and BMP signaling. EmNCSCs expressed several transcription factors characteristic of early NC specification in chick, mouse, and Xenopus. For instance, MSX1 and ZIC1, two of the earliest NC genes, were upregulated 48 and 72 hours, respectively, after the start of neuralization. These two genes have been demonstrated to be pivotal in the development of the NC and the expression of later NC-specific genes , , , . The later NC-specifying genes include Sox9, Sox10 and Snail, which proceed to regulate the EMT and migration of NC cells from the neural tube , , . Consistent with these data, expression of Sox9, Sox10 and SNAIL was upregulated during neuralization of hESCs. The chronological order of upregulation of some NC genes in the hESC cultures may follow a different timetable compared to known in vivo models; whether this is due to an in vitro artifact or reflects species differences remains to be determined.
In vivo, the NC competency zone (area around the neuroectoderm border) is mediated by BMP and Wnt signaling with additional signals, including Wnts mediating the induction of the NC , ,  and delamination of NC from the neural tube , . Different from our approach, most protocols for human NC derivation from ES cells use exogenous BMP to facilitate the differentiation , , . To test if BMP pathway is also important for the generation of emNCSCs, we treated neurosphere cultures with Noggin. After treatment with Noggin, the expression of NC markers Sox10, p75 and HNK-1 was nearly ablated compared to controls. Treatment with the Wnt/β-catenin antagonist Dkk1 resulted in the ablation of Sox10 and HNK1, but not p75. These results are consistent with the distinct mechanisms of action of Noggin and Dkk1. While Noggin (produced by the notochord in vivo) re-specifies the dorsal neuroepithelium into ventral fates, Dkk1 inhibits the early steps of NC induction, without altering the axial cell fates or general patterning . In our hands, the dorsal neuroepithelium-like clusters (adherent neurspheres) are positive for p75. The loss of Sox10 and p75 after the addition of Noggin is consistent with re-specification of dorsal neuroepithelium into ventral neuroepithelium, which typically does not express NC markers. Following the same logic, the addition of Dkk1 may inhibit the NC specification, i.e. appearance of Sox10- and HNK-1-positive emNCSCs, but does not re-specify the dorsal neuroepithelium-like cells, which remain positive for p75. Indeed, the human neural tube cells were found positive for the p75 antigen at the time when the NC is generated . It remains to be determined if this expression is localized to the premigratory neural crest in the dorsal part of human neural tube.
In addition to the patterning growth factors discussed above, matrix elasticity may affect hESC differentiation towards NC lineages, similar to that seen for muscle differentiation , . It will be important to investigate hESC differentiation into emNCSC using various elasticity matrixes.
Finally, we assessed the ability of emNCSCs to migrate in vivo, incorporate into NC derivatives and differentiate appropriately. Grafted emNCSCs efficiently contributed to a variety of NC-derived tissues and differentiated appropriately. This finding demonstrates that emNCSCs are competent to contributing to NC derivatives such as the trigeminal ganglia, as cells that do not normally incorporate into the trigeminal ganglia are excluded from the developing ganglia even when they are immediately adjacent to it . Furthermore, emNCSCs contributed to connective tissues including smooth muscle. Therefore, emNCSCs are capable of extensive migration to appropriate NC cell destinations and appear to have the ability to interact with adjacent host NC cells and differentiate efficiently compared to late emNCSCs, in vivo . Critically, transplanted human emNCSCs do not contribute to the non-NC cell types, such as CNS tissue (e.g., midbrain).
Given the fact that these cells are derived in an antigen free environment (no co-culture with murine tissue) the protocol for the derivation of emNCSCs is ideal for generating emNCSC-derived tissues in the culture dish that can subsequently be used for patient treatment (e.g., neurocristopahties). To determine the therapeutic potential of these cells for treating a disease model, we investigated the ability of emNCSCs to colonize aganglionic embryonic guts in organotypic cultures. EmNCSCs were found to be capable of colonizing aganglionic guts and differentiating into neurofilament-positive cells, presumably enteric neurons, in ex vivo gut cultures. This suggests that emNCSCs might be useful in cell replacement therapies to treat neurocristopathies.
Molecular characterization of hESC neuralization. (A–D) Microarray profiles of hESC derived neruospheres over a 10-day (240 hour) time period. Markers are indicated in the figure. Y-axis shows the fold increase, X-axis shows the time (hours). (E) The Q-PCR profile for Pax3 induction, X-axis shows hours of induction (0-144), Y-axis indicates fold increase of Pax3.
(0.12 MB TIF)
Spontaneously differentiated emNCSC cultures. Spontaneously differentiated emNCSC cultures yield a mix of cell types: (A) SMA (green), Peripherin (red), (B) GFAP (green), and MAP2 (red) positive cells, (C) GFAP (red), and HNK1 (green) positive cells.
(0.70 MB TIF)
The authors would like to thank A. Miller and J. Lesperance for the help with hESC cultures, R. Bajpai for RNA preparation, Y. Altman for assistance with FACS sorting, K. Liu for microarray and Q-PCR assistance, L. Creighton for help with the immunhistochemistry on chondrocytes, D. Arce and M. Jones for assistance with cryosectioning, M. Wegner and M. Fukuda for provision of antibodies, R. Gupta and M. Mercola for providing the DKK plasmid, T. Cheung for DKK media preparation, E. Chavez and A. Ryan for providing timed pregnant Ret.k- mice.
Conceived and designed the experiments: MBF AVT. Performed the experiments: CLC JM SM GC FC MN. Analyzed the data: SM ES MBF AVT. Contributed reagents/materials/analysis tools: ES MBF AVT. Wrote the paper: CLC JM AVT.
- 1. Morrison SJ, White PM, Zock C, Anderson DJ (1999) Prospective identification, isolation by flow cytometry, and in vivo self-renewal of multipotent mammalian neural crest stem cells. Cell 96: 737–749.
- 2. Wong CE, Paratore C, Dours-Zimmermann MT, Rochat A, Pietri T, et al. (2006) Neural crest-derived cells with stem cell features can be traced back to multiple lineages in the adult skin. J Cell Biol 175: 1005–1015.
- 3. Yoshida S, Shimmura S, Nagoshi N, Fukuda K, Matsuzaki Y, et al. (2006) Isolation of multipotent neural crest-derived stem cells from the adult mouse cornea. Stem Cells 24: 2714–2722.
- 4. Kawaguchi J, Nichols J, Gierl MS, Faial T, Smith A (2010) Isolation and propagation of enteric neural crest progenitor cells from mouse embryonic stem cells and embryos. Development 137: 693–704.
- 5. LaBonne C, Bronner-Fraser M (1999) Molecular mechanisms of neural crest formation. Annu Rev Cell Dev Biol 15: 81–112.
- 6. LaBonne C, Bronner-Fraser M (1998) Neural crest induction in Xenopus: evidence for a two-signal model. Development 125: 2403–2414.
- 7. Newgreen DF, Gooday D (1985) Control of the onset of migration of neural crest cells in avian embryos. Role of Ca++-dependent cell adhesions. Cell Tissue Res 239: 329–336.
- 8. O'Rahilly R, Muller F (2007) The development of the neural crest in the human. J Anat 211: 335–351.
- 9. Yu HS (2002) Melanocyte destruction and repigmentation in vitiligo: a model for nerve cell damage and regrowth. J Biomed Sci 9: 564–573.
- 10. Tagra S, Talwar AK, Walia RL, Sidhu P (2006) Waardenburg syndrome. Indian J Dermatol Venereol Leprol 72: 326.
- 11. Marsot-Dupuch K, Smoker WR, Grauer W (2004) A rare expression of neural crest disorders: an intrasphenoidal development of the anterior pituitary gland. AJNR Am J Neuroradiol 25: 285–288.
- 12. Mizuseki K, Sakamoto T, Watanabe K, Muguruma K, Ikeya M, et al. (2003) Generation of neural crest-derived peripheral neurons and floor plate cells from mouse and primate embryonic stem cells. Proc Natl Acad Sci U S A 100: 5828–5833.
- 13. Pomp O, Brokhman I, Ben-Dor I, Reubinoff BE, Goldstein RS (2005) Generation of peripheral sensory and sympathetic neurons and neural crest cells from human embryonic stem cells. Stem Cells.
- 14. Lazzari G, Colleoni S, Giannelli SG, Brunetti D, Colombo E, et al. (2006) Direct derivation of neural rosettes from cloned bovine blastocysts: a model of early neurulation events and neural crest specification in vitro. Stem Cells 24: 2514–2521.
- 15. Lee G, Kim H, Elkabetz Y, Al Shamy G, Panagiotakos G, et al. (2007) Isolation and directed differentiation of neural crest stem cells derived from human embryonic stem cells. Nat Biotechnol 25: 1468–1475.
- 16. Jiang X, Gwye Y, McKeown SJ, Bronner-Fraser M, Lutzko C, et al. (2008) Isolation and characterization of neural crest stem cells derived from in vitro differentiated human embryonic stem cells. Stem Cells Dev.
- 17. Chambers SM, Fasano CA, Papapetrou EP, Tomishima M, Sadelain M, et al. (2009) Highly efficient neural conversion of human ES and iPS cells by dual inhibition of SMAD signaling. Nat Biotechnol 27: 275–280.
- 18. Colleoni S, Galli C, Giannelli SG, Armentero MT, Blandini F, et al. (2010) Long-term culture and differentiation of CNS precursors derived from anterior human neural rosettes following exposure to ventralizing factors. Exp Cell Res 316: 1148–1158.
- 19. Basch ML, Bronner-Fraser M, Garcia-Castro MI (2006) Specification of the neural crest occurs during gastrulation and requires Pax7. Nature 441: 218–222.
- 20. Kleber M, Lee HY, Wurdak H, Buchstaller J, Riccomagno MM, et al. (2005) Neural crest stem cell maintenance by combinatorial Wnt and BMP signaling. J Cell Biol 169: 309–320.
- 21. Bronner-Fraser M (2002) Molecular analysis of neural crest formation. J Physiol Paris 96: 3–8.
- 22. Paratore C, Goerich DE, Suter U, Wegner M, Sommer L (2001) Survival and glial fate acquisition of neural crest cells are regulated by an interplay between the transcription factor Sox10 and extrinsic combinatorial signaling. Development 128: 3949–3961.
- 23. Selleck MA, Garcia-Castro MI, Artinger KB, Bronner-Fraser M (1998) Effects of Shh and Noggin on neural crest formation demonstrate that BMP is required in the neural tube but not ectoderm. Development 125: 4919–4930.
- 24. Bajpai R, Coppola G, Kaul M, Talantova M, Cimadamore F, et al. (2009) Molecular stages of rapid and uniform neuralization of human embryonic stem cells. Cell Death Differ 16: 807–825.
- 25. Taneyhill LA, Bronner-Fraser M (2005) Dynamic alterations in gene expression after Wnt-mediated induction of avian neural crest. Mol Biol Cell 16: 5283–5293.
- 26. Rao MS, Anderson DJ (1997) Immortalization and controlled in vitro differentiation of murine multipotent neural crest stem cells. J Neurobiol 32: 722–746.
- 27. Maurer J, Fuchs S, Jager R, Kurz B, Sommer L, et al. (2007) Establishment and controlled differentiation of neural crest stem cell lines using conditional transgenesis. Differentiation 75: 580–591.
- 28. Shah NM, Groves AK, Anderson DJ (1996) Alternative neural crest cell fates are instructively promoted by TGFbeta superfamily members. Cell 85: 331–343.
- 29. Ido A, Ito K (2006) Expression of chondrogenic potential of mouse trunk neural crest cells by FGF2 treatment. Dev Dyn 235: 361–367.
- 30. Durbec PL, Larsson-Blomberg LB, Schuchardt A, Costantini F, Pachnis V (1996) Common origin and developmental dependence on c-ret of subsets of enteric and sympathetic neuroblasts. Development 122: 349–358.
- 31. Hearn CJ, Young HM, Ciampoli D, Lomax AE, Newgreen D (1999) Catenary cultures of embryonic gastrointestinal tract support organ morphogenesis, motility, neural crest cell migration, and cell differentiation. Dev Dyn 214: 239–247.
- 32. Lang D, Chen F, Milewski R, Li J, Lu MM, et al. (2000) Pax3 is required for enteric ganglia formation and functions with Sox10 to modulate expression of c-ret. J Clin Invest 106: 963–971.
- 33. Serbedzija GN, Bronner-Fraser M, Fraser SE (1989) A vital dye analysis of the timing and pathways of avian trunk neural crest cell migration. Development 106: 809–816.
- 34. Wakamatsu Y, Weston JA (1997) Sequential expression and role of Hu RNA-binding proteins during neurogenesis. Development 124: 3449–3460.
- 35. Elkabetz Y, Panagiotakos G, Al Shamy G, Socci ND, Tabar V, et al. (2008) Human ES cell-derived neural rosettes reveal a functionally distinct early neural stem cell stage. Genes Dev 22: 152–165.
- 36. Barembaum M, Bronner-Fraser M (2005) Early steps in neural crest specification. Semin Cell Dev Biol 16: 642–646.
- 37. Bronner-Fraser M (2003) Hierarchy of events regulating neural crest induction. Harvey Lect 99: 129–144.
- 38. Henderson DJ, Copp AJ (1997) Role of the extracellular matrix in neural crest cell migration. J Anat 191(Pt 4): 507–515.
- 39. Wilson YM, Richards KL, Ford-Perriss ML, Panthier JJ, Murphy M (2004) Neural crest cell lineage segregation in the mouse neural tube. Development 131: 6153–6162.
- 40. Thomas S, Thomas M, Wincker P, Babarit C, Xu P, et al. (2008) Human neural crest cells display molecular and phenotypic hallmarks of stem cells. Hum Mol Genet 17: 3411–3425.
- 41. Duband JL (2006) Neural crest delamination and migration: integrating regulations of cell interactions, locomotion, survival and fate. Adv Exp Med Biol 589: 45–77.
- 42. Newgreen DF, Minichiello J (1996) Control of epitheliomesenchymal transformation. II. Cross-modulation of cell adhesion and cytoskeletal systems in embryonic neural cells. Dev Biol 176: 300–312.
- 43. Pla P, Moore R, Morali OG, Grille S, Martinozzi S, et al. (2001) Cadherins in neural crest cell development and transformation. J Cell Physiol 189: 121–132.
- 44. Bannerman P, Nichols W, Puhalla S, Oliver T, Berman M, et al. (2000) Early migratory rat neural crest cells express functional gap junctions: evidence that neural crest cell survival requires gap junction function. J Neurosci Res 61: 605–615.
- 45. Bobola N, Engist B (2008) IGFBP5 is a potential regulator of craniofacial skeletogenesis. Genesis 46: 52–59.
- 46. Relaix F, Rocancourt D, Mansouri A, Buckingham M (2004) Divergent functions of murine Pax3 and Pax7 in limb muscle development. Genes Dev 18: 1088–1105.
- 47. Bronner-Fraser M (1986) Analysis of the early stages of trunk neural crest migration in avian embryos using monoclonal antibody HNK-1. Dev Biol 115: 44–55.
- 48. Portier MM, de Nechaud B, Gros F (1983) Peripherin, a new member of the intermediate filament protein family. Dev Neurosci 6: 335–344.
- 49. Anderson DJ, Groves A, Lo L, Ma Q, Rao M, et al. (1997) Cell lineage determination and the control of neuronal identity in the neural crest. Cold Spring Harb Symp Quant Biol 62: 493–504.
- 50. Baroffio A, Dupin E, Le Douarin NM (1988) Clone-forming ability and differentiation potential of migratory neural crest cells. Proc Natl Acad Sci U S A 85: 5325–5329.
- 51. Le Douarin NM, Calloni GW, Dupin E (2008) The stem cells of the neural crest. Cell Cycle 7: 1013–1019.
- 52. Raible DW (2006) Development of the neural crest: achieving specificity in regulatory pathways. Curr Opin Cell Biol 18: 698–703.
- 53. Carmona-Fontaine C, Acuna G, Ellwanger K, Niehrs C, Mayor R (2007) Neural crests are actively precluded from the anterior neural fold by a novel inhibitory mechanism dependent on Dickkopf1 secreted by the prechordal mesoderm. Dev Biol 309: 208–221.
- 54. Creuzet S, Couly G, Le Douarin NM (2005) Patterning the neural crest derivatives during development of the vertebrate head: insights from avian studies. J Anat 207: 447–459.
- 55. Sato T, Sasai N, Sasai Y (2005) Neural crest determination by co-activation of Pax3 and Zic1 genes in Xenopus ectoderm. Development 132: 2355–2363.
- 56. Dottori M, Gross MK, Labosky P, Goulding M (2001) The winged-helix transcription factor Foxd3 suppresses interneuron differentiation and promotes neural crest cell fate. Development 128: 4127–4138.
- 57. Cheung M, Briscoe J (2003) Neural crest development is regulated by the transcription factor Sox9. Development 130: 5681–5693.
- 58. Sander M, Paydar S, Ericson J, Briscoe J, Berber E, et al. (2000) Ventral neural patterning by Nkx homeobox genes: Nkx6.1 controls somatic motor neuron and ventral interneuron fates. Genes Dev 14: 2134–2139.
- 59. Pattyn A, Morin X, Cremer H, Goridis C, Brunet JF (1999) The homeobox gene Phox2b is essential for the development of autonomic neural crest derivatives. Nature 399: 366–370.
- 60. Stanke M, Junghans D, Geissen M, Goridis C, Ernsberger U, et al. (1999) The Phox2 homeodomain proteins are sufficient to promote the development of sympathetic neurons. Development 126: 4087–4094.
- 61. Grillet N, Dubreuil V, Dufour HD, Brunet JF (2003) Dynamic expression of RGS4 in the developing nervous system and regulation by the neural type-specific transcription factor Phox2b. J Neurosci 23: 10613–10621.
- 62. Nikopoulos GN, Duarte M, Kubu CJ, Bellum S, Friesel R, et al. (2007) Soluble Jagged1 attenuates lateral inhibition, allowing for the clonal expansion of neural crest stem cells. Stem Cells 25: 3133–3142.
- 63. Whitlock KE, Wolf CD, Boyce ML (2003) Gonadotropin-releasing hormone (GnRH) cells arise from cranial neural crest and adenohypophyseal regions of the neural plate in the zebrafish, Danio rerio. Dev Biol 257: 140–152.
- 64. Whitlock KE, Smith KM, Kim H, Harden MV (2005) A role for foxd3 and sox10 in the differentiation of gonadotropin-releasing hormone (GnRH) cells in the zebrafish Danio rerio. Development 132: 5491–5502.
- 65. Garcia-Castro MI, Marcelle C, Bronner-Fraser M (2002) Ectodermal Wnt function as a neural crest inducer. Science 297: 848–851.
- 66. Anderson RB, Newgreen DF, Young HM (2006) Neural crest and the development of the enteric nervous system. Adv Exp Med Biol 589: 181–196.
- 67. Marchant L, Linker C, Ruiz P, Guerrero N, Mayor R (1998) The inductive properties of mesoderm suggest that the neural crest cells are specified by a BMP gradient. Dev Biol 198: 319–329.
- 68. Fang D, Leishear K, Nguyen TK, Finko R, Cai K, et al. (2006) Defining the conditions for the generation of melanocytes from human embryonic stem cells. Stem Cells 24: 1668–1677.
- 69. Hong CS, Saint-Jeannet JP (2007) The activity of Pax3 and Zic1 regulates three distinct cell fates at the neural plate border. Mol Biol Cell 18: 2192–2202.
- 70. Tribulo C, Aybar MJ, Nguyen VH, Mullins MC, Mayor R (2003) Regulation of Msx genes by a Bmp gradient is essential for neural crest specification. Development 130: 6441–6452.
- 71. Monsoro-Burq AH, Wang E, Harland R (2005) Msx1 and Pax3 cooperate to mediate FGF8 and WNT signals during Xenopus neural crest induction. Dev Cell 8: 167–178.
- 72. Meulemans D, Bronner-Fraser M (2004) Gene-regulatory interactions in neural crest evolution and development. Dev Cell 7: 291–299.
- 73. Sakai D, Suzuki T, Osumi N, Wakamatsu Y (2006) Cooperative action of Sox9, Snail2 and PKA signaling in early neural crest development. Development 133: 1323–1333.
- 74. Knecht AK, Bronner-Fraser M (2002) Induction of the neural crest: a multigene process. Nat Rev Genet 3: 453–461.
- 75. Burstyn-Cohen T, Stanleigh J, Sela-Donenfeld D, Kalcheim C (2004) Canonical Wnt activity regulates trunk neural crest delamination linking BMP/noggin signaling with G1/S transition. Development 131: 5327–5339.
- 76. Shoval I, Ludwig A, Kalcheim C (2007) Antagonistic roles of full-length N-cadherin and its soluble BMP cleavage product in neural crest delamination. Development 134: 491–501.
- 77. Gossrau G, Thiele J, Konang R, Schmandt T, Brustle O (2007) Bone morphogenetic protein-mediated modulation of lineage diversification during neural differentiation of embryonic stem cells. Stem Cells 25: 939–949.
- 78. Engler AJ, Sweeney HL, Discher DE, Schwarzbauer JE (2007) Extracellular matrix elasticity directs stem cell differentiation. J Musculoskelet Neuronal Interact 7: 335.
- 79. Even-Ram S, Artym V, Yamada KM (2006) Matrix control of stem cell fate. Cell 126: 645–647.
- 80. McKeown SJ, Newgreen DF, Farlie PG (2003) Temporal restriction of migratory and lineage potential in rhombomere 1 and 2 neural crest. Dev Biol 255: 62–76.